Grain-size Modes in Long Island Sands and Silts may be Inherited from
Grain-size Modes in Bedrock
Courtney Melrose* and Dr. Gil Hanson
Stony Brook University
*MS Candidate
Introduction:
During a study of loess-like deposits in Suffolk County, Long Island, New York,
(Figure 1) it was noted that the grain sizes have a distinct set of modes at about 20μm, and
400μm in most if not all loess samples and modes of 20μm, 80 μm and 400μm (Figure 2) in
samples of till and outwash
on the Stony Brook campus
(Zhong, 2002, Kundic,
2005, A. Olaofe, personal
communication, 2012 and
T. Clare, 2013). We found
that the same set of modes
were also present in
Cretaceous
sediments,
weathered, Long Island,
basement rocks and mechanically crushed rocks
typical for Long Island. All
of the samples were ana- Figure 1. Study locations in Suffolk County, Long Island New York.
lyzed with a Malvern Mastersizer 2000.
Loess is windblown silt, predominantly composed of quartz. Deposits of loess display a
lack of bedding, and are usually dun in color. Loess deposits are prevalent on Long Island, and
are the reason for Long Island’s rich agriculture. Mastersizer analysis of loess-like samples collected from Stony Brook showed a pattern of modes around 20μm and 400μm (Figure 2). The
modes at 20 and 400 μm were assumed to be due to processes related to sorting by wind. The
400μm fraction would be a result of travel by saltation from proximal sources and the 20μm
fraction would be a result of travel by suspension from more distal sources. This assumes that
the source of a wind blown sediment has a wide range of grain sizes and a well graded distribution of grain sizes from silt through sand. That is, each grain-size fraction has a near equal proportion of grains by weight or volume in sediment sources. As we will show that is not necessarily the case for the loess that we have studied.
These same loess deposits also contained randomly placed pebbles, which is
uncharacteristic of a wind blown silt or sand. This fact led to these loess deposits being called
“pebbly-loess” or “loess-like” deposits. To investigate if these modes were unique to pebbly
loess and glacial sediments, or whether they were also present in Cretaceous clays, sands and
gravels or weathered granite, samples of Cretaceous sediments were collected at Caumsett State
Park and sediments and weathered basement from cores from Brookhaven National Laboratory
(DeLaguna, 1963). To test if these modes were inherent in the potential parent rock of the loess,
three hand specimens of variably sheared granite gneisses found on the Stony Brook Campus
(Fig. 3) which are similar to the Stony Creek granite in Connecticut were crushed, ground and
pulverized. While not every sample analyzed displayed modes at 20μm, 80 μm
and 400μm, every sample shows a peak at
or near one or more of these modes .
Methods:
A Malvern Mastersizer 2000 uses
laser diffraction to accurately analyze the
grain-size distribution of particles.
Figure 2. Histograms of grain-size for the less than 1 mm
particles in loess and till found at one section on the Stony
Brook University campus. The upper histogram is based on
one phi fractions for 6 samples of loess determined using wet
and dry sieving and settling tube (Zhong, 2005). The loess
has modes at 250 and 20 and less than one micron.
The lower histogram (Clare, 2013) ,(note that the
values on the abscissa are reversed.) was determined using
the Malvern Mastersizer 2000 for loess and till at the same
site as that for the samples in the upper graph. Essentially the
same three modes are shown in the loess and also in the underlying till at 10 to 20 microns and a smaller clay mode at
about 0.5 microns.
“During
the
laser
diffraction
measurement, particles are passed
through a focused laser beam. These
particles scatter light at an angle that is
inversely proportional to their size. The
angular intensity of the scattered light is
then measured by a series of
photosensitive detectors. The number and
positioning of these detectors in the Mastersizer 2000 has been optimized to
achieve maximum resolution across a
broad range of sizes. The map of scattering intensity versus angle is the primary source of information used to calculate
the particle size. The
scattering of
particles is accurately predicted by the
Mie scattering model. This model is rigorously applied within the Mastersizer 2000
software, allowing accurate sizing across
the
widest
possible
dynamic
range.” (http://www.malvern.com/Assets/
MRK501.pdf)
The procedure for analysis is that
of Sperazza et. al. 2004. The samples
were passed through a 1mm sieve to remove the coarser fraction. They were then
left in a solution of 5.5 g/l of sodium hexametaphosphate (Na(PO3)6), (a deffloculant) overnight. For analysis each sample was introduced by pipette into a beaker containing
500ml of the (Na(PO3)6),solution until the obscuration percent readout on the machine was between 15% and 23%.
Four different rock crushing procedures were used to convert the whole rocks from
Stony Brook University campus into “artificial sediments”. We first either broke or cut sections
off of each rock (labeled rock A, rock B, and rock C) (figure 3). These sections were first
weighed and then crushed. These methods were designed to mimic the effects of glacial action
on these rocks. Five set samples of each of the three rocks was produced, and four separate
crushing regimens were used. For the first set, the sections of rocks were crushed by hand in a
metal mortar and pestle. This produced chunky gravel. The second set included a grinding by
hand, in a ceramic mortar and pestle. The third set included both of the previous grinding methods, but the sample was subsequently placed in an agate shatterbox to pulverize the sample.
Each rock sample (rock A, B, and C), was processed this way for approximately two minutes.
For the fourth set, the samples were processed
in a metal grinding mill.
Locations:
Deposits of fine quartz rich silt, or loess, containing pebbles were sampled at Stony
Brook where this loess overlies a one meter
thick glacial till. According to Zhong, 2001,
40
Ar/39Ar muscovite in the loess gives ages
dominantly between 200 and 450 Ma which Figure 3. The three specimens of granite gneisses colsuggests that the muscovite is derived from lected on the Stony Brook campus, used for whole rock
basement rocks in New England, consistent
with the direction of glacial advance to the
south. OSL dating for this loess gives an age for deposition of 13 to 14 Ka (Kundic, 2005) long
after the last glacier retreated from Long Island at about 20 Ka. This loess was deposited at
about the beginning of the Younger
Dryas when the climate returned to near
Glacial Maximum conditions with lasted
for some 1500 years..
Samples were collected at
Caumsett State Park from both glacial
(Pleistocene age) and Cretaceous sediments. Muscovite in the glacial and
Cretaceous sediments give 40Ar/39Ar
ages between 200 and 400 Ma again
suggesting that the muscovite is derived
from basement rocks in New England
(Zhong, 2001)
Figure 4. Histogram showing the grain-size modes for the whole
Samples from Brookhaven Narock first crushing series for Rock A.
tional Laboratory were collected from
two separate cores, S6409 and S6434 (De Laguna, 1963). Each sample was chosen for its sediment type, and location in depth according to Figure 3 of De Laguna, 1963. A total of 17 samples were analyzed, of these, five were from the Magothy formation, two were from the Raritan
Clay, four were from the Lloyd sand and six were of weathered bedrock.
Results and Discussion:
The histograms show modes based on the volume percent of sediment of a particular
size in the Mastersizer machine. The actual modes vary slightly between samples, but generally
occur around 20μm, 80μm, and 400μm. The Cretaceous samples from Caumsett State Park in
Fig. 9 illustrate this clustering.
One hypothesis that we developed before we analyzed the Cretaceous samples was that
the modes in the glacial sediments were produced by the grinding action of the glacier. The
same modes at about 20µm, 80µm, and 400µm were found in the whole rock crushing experiments (Fig. 4). This would support the hypothesis that the modes might either be inherent in the
rocks, or are produced by the grinding by the glacier.
Each sample from Caumsett State Park, whether from glacial or Cretaceous sediments,
has peaks at around 20μm, 80μm, and 400μm (figures 5 through 9). This would indicate that
these modes could not have been produced by grinding by a glacier but may actually be
inherent in the source bedrocks of these clastic sediments.
This is further supported by the findings obtained for samples from the cores at
Brookhaven National Laboratory. Figure 10 shows modes at 20, 65 and 500m for the
weathered remains of granite.
These results suggest that for this set of Pleistocene and Cretaceous sediments, based on
the weathering products of granite and crushed granite, the grain-size modes found in these
clastic sediments are inherent in the bedrock source and are inherited by the sediments derived
from such bedrock.
Acknowledgments:
We thank Sean Tvelia for his help, and the use of his field equipment, Jazmin Inoa for
her assistance at the beginning of this project, Michael Sperazza and Kara Dias for instruction
and the use of the Malvern Mastersizer 2000, Steve Melrose for designing and creating the jigs
used for cutting open sample cores and helping to collect samples and Michael Noll for help in
accessing the samples from the Brookhaven National Laboratory bore hole cores.
References Cited:
Clare, T., 2013. Grain Size Analysis of Loess and Glacial Sediments at Stony Brook University,
https://dl.dropboxusercontent.com/u/14127214/Clare%20Grain%20Size%20Analysis%
20of%20Glacial%20Sediments%20at.docx
De Laguna, W., 1963. Geology of Brookhaven National Laboratory and Vicinity, Suffolk
County New York. Geological Survey Bulletin 1156-A.
http://pubs.er.usgs.gov/publication/b1156A
Kundic, V. 2005. Age and Provenance of Long Island Loess. MS thesis, Stony Brook University, 61 p.
http://www.geo.sunysb.edu/reports/kundic-thesis.pdf
Sperazza, M. and Reimer, P.J., and Hendrix, M.S. 2004. High Resolution Particle Size Analysis
of Naturally Occurring Very Fine Grained Sediment Through Laser Diffractometry. Journal
of Sedimentary Research. Vol. 74 pg736-743.
Zhong, J. 2001. Evaluation of Ar-Ar ages of Individual Mica Grains for Provenance Studies of
Loess, Long Island, NY. MS thesis, Stony Brook University, 35 p.
http://www.geo.sunysb.edu/reports/zhong-MS-thesis.pdf
Figure 5. Histogram of glacial sand from Caumsett
State Park with modes at 100μm and 400μm.
Figure 7. Histogram of glacial sand from Caumsett
State Park with modes at around 100μm, and 400μm.
Figure 9. Histogram of the Cretaceous samples from
Caumsett State Park, showing the particle size variation
between samples, with a clustering at about 80μm and
400μm.
Figure 6. Histogram of Cretaceous sand from Caumsett
State Park, with modes around 80μm and 300μm, with a
small peak possible at 20μm.
Figure 8. Histogram of Cretaceous silty sand from
Caumsett State Park, with peaks around 20μm, 60μm
and 300μm
Figure 10. Histogram of the product of weathered
granite from USGS core S6409 at a depth of 1501 to
1565 feet which shows the same three modes at about
20 , 65 and 500.